At Mon, 21 Jan 2013 11:47:40 +0900,
Hitoshi Mitake wrote:
> +
> +# forbid empty lines after break; line of switch statement
> +# e.g.
> +# + break;
> +# +
> +# + case XXX:
> + if ($line =~ /\bbreak;/) {
> + my ($nlength, $nindent) = line_stats($line);
> +
> + my $ln = $linenr + 1;
> + while ($lines[$ln] =~ /^\+$/) { $ln++; }
> + my $sline = $lines[$ln];
> +
> + if ($sline =~ /\bcase/ || $sline =~ /\bdefault/) {
/\b(case|default)/ looks simpler.
> + my ($snlength, $snindent) = line_stats($sline);
> + if ($nindent - 8 == $snindent) {
($nindent > $snindent) looks better to catch more errors.
> + ERROR("NL_AFTER_BREAK_IN_SWITCH",
> + "don't insert empty lines after break; line
> of switch statement" . $herecurr);
Add "\n" to the end of the second argument.
In addition, you patch allows the following. Is it expected behavior?
+ switch (x) {
+ case A:
+ break;
+ default:
+ break; /* shouldn't this be an error? */
+
+ }
Thanks,
Kazutaka
--
sheepdog mailing list
sheepdog@lists.wpkg.org
http://lists.wpkg.org/mailman/listinfo/sheepdog
